Effective Macro Placement for Very Large Scale Designs Using MCTS Guided by Pre-Trained RL
Jai-Ming Lin, Zong-Ze Lee, Nan-Chu Lin · 2025
Macro placement plays a critical role in modern designs. Some researchers have applied reinforcement learning (RL) techniques to handle this problem. This paper proposes an effective placer based on the Monte Carlo Tree Search (MCTS) algorithm, guided by a pretrained RL agent. To reduce the complexities of RL and MCTS, we transform the macro placement problem into a macro group allocation problem. Additionally, we propose a new reward function to facilitate training convergence in RL. Moreover, to reduce runtime without affecting placement quality, we use the pretraining result to directly evaluate the placement quality in MCTS. Experiments show that our MCTS-based placer can achieve high-quality results even in the early stages of RL training. Moreover, our method outperforms state-of-the-art placers.